Resnet10

🚀 JAX to PyTorch Converted Model (ResNet-10)

This project involves converting a JAX model to PyTorch, specifically a ResNet - 10 model. It aims to make the model more accessible for robotics tasks due to its small size and preserve the original weights during the conversion.

🚀 Quick Start

from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained("helper2424/resnet10")

✨ Features

• Converted from JAX to PyTorch, making it available for PyTorch users.
• Preserves the original model's architecture and weights.
• Small - sized ResNet - 10 model, suitable for robotics tasks.

📦 Installation

No specific installation steps are provided in the original README.

💻 Usage Examples

Basic Usage

from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained("helper2424/resnet10")

📚 Documentation

Model Description

This model is a PyTorch port of the original JAX implementation. The conversion maintains the original model's architecture and weights while making it accessible to PyTorch users. The original model is from https://github.com/rail-berkeley/hil-serl/blob/7d17d13560d85abffbd45facec17c4f9189c29c0/serl_launcher/serl_launcher/utils/train_utils.py#L103.

Model Details

Conversion Process

This model was converted using an automated JAX to PyTorch conversion pipeline, ensuring:

• Weight preservation
• Architecture matching
• Numerical stability

Code

The code for this model can be found at: https://github.com/helper2424/resnet10

📄 License

This project is licensed under the Apache 2.0 license.

📚 Citation

@misc{resnet10,
   title = "Resnet10",
   author = "Eugene Mironov and Khalil Meftah and Adil Zouitine and Michel Aractingi and Ke Wang",
   month = jan,
   year = "2025",
   address = "Online",
   publisher = "Hugging Face",
   url = "https://huggingface.co/helper2424/resnet10",
}